fix(vector_store): normalize QdrantStore.search_vectors() to return "metadata"

QdrantStore.search_vectors() returned results keyed by "payload" while
HybridSearch and PineconeStore both expect/return "metadata". This silently
dropped metadata from Qdrant results and caused HybridSearch.filter_by_metadata
to reject every candidate when a filter was applied (empty result sets).

Fixes #840

semantica/vector_store/qdrant_store.py

@@ -165,7 +165,7 @@ class QdrantCollection:
                     {
                         "id": result.id,
                         "score": result.score,
-                        "payload": result.payload or {},
+                        "metadata": result.payload or {},
                     }
                 )
 

tests/vector_store/test_vector_store_deepdive.py

@@ -232,6 +232,16 @@ class TestVectorStoreDeepDive(unittest.TestCase):
         self.assertEqual(len(results), 1)
         self.assertEqual(results[0]["id"], "vec_1")
 
+        self.assertIn("metadata", results[0])
+        self.assertNotIn("payload", results[0])
+        self.assertEqual(results[0]["metadata"], {"type": "a"})
+
+        hybrid = HybridSearch()
+        matched = hybrid.filter_by_metadata(
+            results, MetadataFilter().eq("type", "a")
+        )
+        self.assertEqual(len(matched), 1)
+
     @patch('semantica.vector_store.weaviate_store.weaviate')
     @patch('semantica.vector_store.weaviate_store.MetadataQuery')
     @patch('semantica.vector_store.weaviate_store.WEAVIATE_AVAILABLE', True)